There may be many reasons why the AF test results are not as expected.
There are usually the following common reasons:
1) Differences in the methodological aspects of the technical platform. The AF value and error range presented by COA are based on the results of the absolute quantitative calculation of digital PCR.
When the standard is detected by other methodologies (such as NGS), the deviation of AF may be amplified. This amplification is related to the type of mutation detected, the copy number and sequence specificity of the target gene, the chromosome ploidy of the sample, the specific experimental technical route adopted, and the subsequent bioinformatics analysis method. Specifically, 1) When the mutation is SNV, the AF value obtained by digital PCR and NGS generally has a small deviation, while the AF value deviation of CNV and fusion mutations is large. From an experimental perspective, digital PCR is an amplification and quantification of the complete gDNA sample, while the sample often needs to be sheared during the NGS library construction process. Shearing usually makes it easy for the AF of CNV and fusion mutations to deviate from that before the shearing. We often observe this phenomenon when using digital PCR to detect CNV and fusion mutation samples before and after the shearing. We have also found this difference after the shearing in some samples with abnormal overall chromosome ploidy. In addition, from the perspective of post-test bioinformatics analysis, SNV The calculation and analysis methods for AF values are relatively mature and unified, and are closer to the calculation principles of digital PCR AF values. Therefore, the differences between NGS and digital PCR results are often small. However, the AF calculation of CNV and fusion mutations is more dependent on bioinformatics algorithms. The schemes are less unified and mature, and the algorithms vary greatly from one company to another. This is one of the reasons for the large difference between it and digital PCR calibration AF.
Furthermore, from a methodological perspective, digital PCR is a simple, single-shot PCR method for absolute quantification. NGS methods, whether using amplicon or capture schemes, may involve multiple or multiplex PCR during library construction and sequencing. This PCR process can produce some biased amplification, leading to deviations in AF values. We have also verified through digital PCR experiments that AF values for PCR products vary between pre- and post-library construction for certain samples and loci. Furthermore, insufficient or uneven NGS sequencing depth can also cause AF values to differ from expectations. For example, at sites with high or low GC content, the sequencing depth may be uneven compared to other regions, leading to AF deviations. Furthermore, at sites with high copy number amplification, localized insufficient sequencing depth can result in the upper limit of detection being lower than the true copy number, necessitating a higher sequencing depth to reflect the true copy number.
These are just some common causes of AF deviation. Overall, the causes of AF discrepancies are multifaceted and complex, requiring specific analysis of each case. When encountering actual cases, we may require the client's original data and experimental details to facilitate analysis.